THE CRYSTAL STRUCTURE OF BARITE - Journal of the American

THE CRYSTAL STRUCTURE OF BARITE. Linus Pauling, and Paul H. Emmett. J. Am. Chem. Soc. , 1925, 47 (4), pp 1026–1030. DOI: 10.1021/ja01681a016...
0 downloads 0 Views 328KB Size
1026

LINUS PAULING AND PAUL H. EMMETT

Vol. 47

separation of potassium and sodium perchlorates, two extractions with intermediate solution of potassium perchlorate being required; ( 2 ) the extraction separation of potassium and lithium perchlorates, one extraction only being required; and (3) the separation of sodium and lithium subsequent to their extraction from potassium, following the procedure of Willard and Smith2 in which sodium chloride is precipitated from solution in M butyl alcohol, by a hydrogen chloride solution in the same solvent. The sodium was determined volumetrically by Mohr’s method. URBANA,ILLINOIS [CONTRIBUTION FROM THE

GATESCHEMICAL LABORATORY, CALIBORNIA TECHNOLOGY, No. 571

INSTITUTE OF

THE CRYSTAL STRUCTURE OF BARITE BY LINUSPAULING AND PAULH. EMMETT RECEIVED

FEBRUARY 10, 1925

PUBLISHED APRIL 4, 1926

Introduction Barite, BaS04, a member of a large class of isomorphous crystals, is described by Grothl as orthorhombic bipyramidal, with axial ratios a:b:c = 0.8152:1:1.3136. This crystal has been recently investigated by Allison2 by means of X-rays. The axial ratios were redetermined by goniometric measurements, using X-rays instead of visible light, and the values 0.8148:1:1.3131 were obtained. Spectrometric measurements were made from thirteen planes, using a tungsten tube operated at 90,000 volts; and values of d/n were then calculated from the angles of reflection of the tungsten lines and the barium K absorption edge. The conclusions drawn from the data as to the structure of the crystal seem, however, to be clearly erroneous. For the author bases the determination of the unit of structure upon the assumption that the unit must have the crystallographic axial ratios, which in fact result from an arbitrary choice of a parametral plane from among the many permitted by the crystallographic data. Moreover, his derived unit of structure is inconsistent with his data, in that he gives in Table IV for plane (102) dcalcd,= 2.792 anddobs. = 5.562; yet his observation shows that d / n = 5.562; and this obviously requires that d = 5.562 n, where n is an integer, and hence does not permit that d = l / Z X 5.562 as calculated (approximately) from his unit of s t r ~ c t u r e . The ~ Groth, “Chemische Krystallographie,” Engelmann, Leipzig, 1908, Vol. 11, p. 388. S. K. Allison, Am. J . Sci., [ 5 ]8,261 (1924). * The author is aware that there is a difficulty here, for he states that this result from (102) may be due to some “accidental” arrangement of atoms; however, it is not possible with any arrangement of atoms in his unit to explain the observed reflection. The analogy suggested to (100) of succinic acid and (010) of succinic anhydride [Yardley, Proc. Roy. Sac., 105A, 461 (1924)l is not pertinent, for in these cases doslad. is twice d/%&a., not one-half of it.

AprilJ 925

CRYSTAL STRUCTURE OF BARITE

1027

author’s conclusion as to the space group involved is also invalidated, since it is based on his unit of structure. It has seemed to us desirable, therefore, to investigate anew the crystal structure of this substance, utilizing data obtained from Laue as well a5 from spectral photographs. This research was made with the help of a grant made to Professor A. A. Noyes by the Carnegie Institution of Washington, for which we wish t o express our appreciation. The Experimental Method Spectral photographs were taken of the K-radiation of molybdenum reflected from the faces (loo), (010) and (001) of a natural crystal. These faces were previously roughened by grinding with fine carborundum .and oil; the face (loo), which was not developed, was ground normal to the other two pinacoids. Reflections from (100) of calcite were taken a t the same time for calibration. The data are given in Table I. The lines have the following wave lengths in Angstrom units: molybdenum Kal, 0.7078; a2,0.7121; p, 0.6311; 0.6197. TAULB I SPECTRAL PKOTOGRAPWIC DATARROM BARITE (W (100)

Order of reflection

n n n n

Line

Angle of reflection

d h

A.

Estimated intensityo

4O 1.0’ 4.429 vw 4.411 m 4 6.0 4.425 S ff1 4 35.4 4.422 ms 012 4 37.0 4.420 vvw 2n Y 8 3.5 vw 8 12.7 4.419 212 P 2n ff1 9 12.5 4.425 mw 212 ff2 9 16.4 4.420 W vvw 12 22.4 4.414 3n P 312 011 13 52.5 4.429 mw 3n ff2 13 58.7 4.426 W (010) n P 6 41 2,715 vvw n 011 7 30.1 2.712 S n 012 7 31.7 2.718 ms 13 27.4 2.712 mw 212 B 2n ff1 15 6 . 5 2.717 s 2% 012 15 12.5 2.715 ms (001) n Y 4 59.8 3.557 VVW n P 5 5.0 3.562 mw n ff1 5 42.0 3.564 S n 012 5 44.2 3.561 S 10 2 . 3 3.555 vvw 2n Y 3.562 W 2% P 10 12.3 2n ffl 11 27.9 3,560 m 2n 012 11 32.1 3.551 W a The abbreviations signify: s, strong: ms, medium strong; m, medium; mw, medium weak; w, weak; vw, very weak; vvw, very very weak. Y

P

1028

Vol. 47

LINUS PAWING AND PAUL H. EMMETT

Laue photographs4 were made with the incident beam a t small angles with the normal to (OOl), the crystal slips used being cleaved along (001) from the large specimen. The radiation had a minimum wave length of 0.24 8., for the tungsten tube used was operated at a peak voltage of about 52,000 volts. Gnomonic projections were used in interpreting the photographs. Laue data are given in Tables I1 and 111. L A U E PHOTOGRAPHIC

h kl

131

dhii

nX

A.

Estimated

A.

intensity”

1.407 0.42 6’11 1.400 .a4 331 1.285 .29 341 1.285 .42 441 1.199 .42 251 1.114 .42 251 1.114 .33 251 1.114 .30 731 1.048 .39 641 1.025 .42 732 1.016 .41 a signifies absent.

-

TABI,EI1 DATAPROM BARITS:PHOTOGRAPH 1

a 5 0.02 1.6 1.6 3 0.1 .05 .2 .3 1.6

nh

Estimated

A.

A.

intensity

0.960 .960 .931 .881 .802 .793 .785 .770 ,735 ,728 -664‘

0.40 .38 .42 .30 .42 .41 .42 .42 .40 .41 .34

dhkl

hkl

911 -

911 -162 461 563 942 -

10.5.2

943 9 -52 11.3.2 13.1.2

-

,

0.4 .4 1.3 0.01 1.0 0.6 .1 .05 .1 .8 .05

The Unit of Structure Spectral data in Table I give for d/n for (loo), (010) and (OOl), values of 4.423, 2.715 and 3.55 8. On assigning indices on the basis of the unit obtained by putting n = 1 in each case and calculating values of nX, a number of planes such as (%1), (gl), (q61), (=.1.2), etc., gave values less than 0.24 i%., eliminating this unit, and all units in which n = 1 for one or two of the pinacoids. The smallest unit which accounts for these observed reflections is obtained by taking n = 2 for each of the pinacoids; this unit has dloo = 8.846, dola = 5.430, and do01 = 7.10 8. This unit accounts completely for all Laue data obtained, and may accordingly be accepted as the correct one. The amount of data accounted for is shown by the fact that on one Laue photograph over 200 planes, belonging to 90 diff ereiit forms, gave first-order reflections. Representative data are given in Table 11, in which indices refer to the axes of this unit. It is to be noted that these indices may be obtained from those based on the usual crystallographic axes by doubling h. The number of molecules in this unit is four, corresponding to the density 4.51, which is in good agreement with the directly determined valuesl4.48 to 4.50. The density calculated from the values of d / n obtained A reproduction of a Laue photograph through (001) of barite has been published by Wyckoff, “The Structure of Crystals,” The Chemical Catalog Co., New York, 1924, p. 110.

April, 1925

1029

CRYSTAL STRUCTURE: OF BARITE:

spectrometrically by Allison2 is 4.43. We have estimated the probable error in our spectrographic observations to be less than 0.1% ; the agreement in calculated density and the directly determined values (which in general tend to be low) verifies this estimate.6 The Space Lattice and the Space Group Observed first-order reflections from planes with one or two indices even, with the sum of all three indices even, and with the sum of any two indices even (Table 11) require6 that the lattice underlyiiig the structure be the simple orthorhombic lattice I’o. The types of prism planes giving first-order reflections (Table 111) are such as to eliminate definitely all of the holohedral space groups’ V i to Vk6 (2Di- 1 to 2Di- 16) based on this lattice except Vi, Vi, Vi3 and Vi6. L A U E PHOTOGRAPHIC

1 -

-Photograph hkl

310 310

OZl 051 902 10.0.3 072 11.0.2 11.0.3

dhkl

nX

A.

A.

2.630 2.070 1.430 1.150 0.994 ,829 .811 .784 ,733

0.38 .32 .45 .40 .39 -46 .38 .46 .37

TABLE 111 DATAFROM BARITE. -----Photograph

Estimated intensity

a 2.5 a 0.4 1.0 0.02 a a 1.0

PRISM PLANES

2dhhlcl

nX

hkl

A.

A.

210 120 130 140 701 052 801 061 074 11.0.3

3.520 2.770 1.896 1.439 1.245

0.34 .36 .38 .29 .38 .44 .39 -43 .46 -43

1.110 1.094 0.962 .753 .733

Estimated intensity

6.0 a a

a

0.05 a 0.2 a a

1.5

The observed data show that all types of planes with k = 0 give firstorder reflections, those with h = 0 do so only if k 1 is even, and those with 1 = 0 only if h is even. These data render space groups Vi,Vt and Vk3 improbable, for these space groups require the absence of first-order reflections from certain planes in no prism zone, or in only one, and accordingly do not account for all the observed absences. Space group Vi6 permits first-order reflections from all types of planes with h = 0, from those I is even, and from those with 1 = 0 only if k is with k = 0 only if h even. This space group, which completely explains the experimental observations, is accepted as correct. It is to be noted that to change from indices (hkl) used in this paper to those (hkl)sc based on the assignment

+

+

The agreement between daslod. and dab#. given in Table IV of the previous research is such as to indicate a probable error somewhat greater than 0.1% in the spectrometric observations reported. Pauling and Dickinson, THISJOURNAL, 46, 1615 (1924). R. W. G. Wyckoff, “The Analytical Expression of the Results of the Theory of Space-Groups,’’ Carnegie Inst. Pub., No. 318, 1922.

1030

LINUS PAULING AND PAUL H. EMMETT

VOl. 47

of axes to ITA in 6the course of the development of the theory of space groups, the transformation h = k S G , k = hSG,1 = JsG is required. A complete determination of the atomic arrangement is impracticable a t this time, because of the number of parameters involved. An attempt was made to determine the location of barium atoms alone, using the statistical treatment of intensities successfully applied to uranyl nitrate hexahydrate.6 This treatment should be applicable, since the structure factors for barite are made up mainly of the barium term, on account of the large relative reflecting power of this heavy atom. The possible arrangements of the four barium atoms in the unit are' (%)

(b) (c)

$to

310 3 3 1 1 3 > z 4 944zril46;

130; 4 4 2 , 4 4 2 r 4 4 r 4 4 ouv, ~ 4 - - u o , o u + ~ ~ - v , ~ i i v + ~ . 1 1 1 3 1 1 340

For arrangements (a) and (b) the structure factor in the first order is 4 s for planes with hSG even and kSG 1SG even, and 0 for all other planes. These arrangements are definitely eliminated by the experimental data; for example, (&1)SG is absent, and ( 5 2 l ) s G , with smaller interplanar distance, reflects very strongly a t the same wave length. Such wide discrepancies cannot be explained as due to the effect of sulfur and oxygen atoms. The barium atoms are, therefore, located as in (c). Because of the presence of other atoms no attempt was made to determine the two parameters involved.

+

Summary It is shown by Laue and spectral photographs, interpreted with the aid of the theory of space-groups, that the unit of structure of the orthorhombic crystal barite (BaX04) has dloa = 8.846, do10 = 5.430 and do01 = 7.10 A., and contains 4BaS04. These conclusions are not in accord with those previously drawn by Allison. The data further require that the structure be based on the simple orthorhombic lattice Yo. A consideration of prism reflections led to the assignment of this crystal to space group Vi6. The general arrangement of the four barium atoms was determined, but the two parameters defining their exact locations were not evaluated. PASADBNA, CALIPORNIA